LingFrame

Linguistic Atomization Framework: a computational rhetoric toolkit.

LingFrame splits a text into a five-level hierarchy (theme, paragraph, sentence, word, letter) and runs pluggable analysis modules over it: a heuristic rhetorical evaluation, semantic (TF-IDF) theme networks, verb-tense flow, sentiment and entity extraction. It produces interactive HTML reports and scholarly exports (LaTeX, TEI, CoNLL). Scores are marker-density indicators, not a measure of writing quality.

Try it

git clone https://github.com/4444J99/linguistic-atomization-framework.git
cd linguistic-atomization-framework
python3 -m venv .venv && . .venv/bin/activate
pip install .
lingframe quick tests/fixtures/sample_rhetoric.txt
lingframe analyze tests/fixtures/sample_rhetoric.txt -o report.html

Links

• Source code and README
• Documentation (theory, methodology, limitations, tutorials)
• Corpus provenance and licence notices
• Issues

Licensing

LingFrame's code is MIT-licensed. The texts in the repository's corpus/ directory keep their own terms; most are US public domain, but some are non-commercial or no-modification licences. See the corpus notice above.